Tracking the growth and frequency of open source startups has been a long-running project at TechCrunch. This column joined the fun in the last few years, noting what seemed to be a rising wave of startups building open source projects that they later monetized.
BuildBuddy built its service to work with Bazel, an open source version of the Google developer tool Blaze, to pick an example from our coverage. Airbyte built its own open source code that it is monetizing, to highlight another. The trend of startups building open source code, and then a business atop it, or doing the latter while contributing to an extant open code base, is now common enough that we’ve published essays on TechCrunch+ strictly dealing with how to build open source startups. Hell, the topic is even cropping up in crypto circles lately.
The Exchange explores startups, markets and money.
Read it every morning on TechCrunch+ or get The Exchange newsletter every Saturday.
Software companies predicated on open source code have in fact become so normalized today that we’ve seen a few companies go public after building with the model. Confluent was one. Hashicorp was another.
Newly released data indicates that venture investors, buoyed by the success of those IPOs and the implied market traction of commercial companies working with open source code, have been liberal with their deal allocation when it comes to backing the upstarts in recent quarters.
The open source model was doing well before the stock market shifted, the economy softened and we entered into a new normal for startup fundraising metrics and performance expectations — less capital available, higher standards for investment and an increased focus on profitability in conjunction with growth. But there’s reason to think that it may do even better in the coming years.
To understand why, we have to ask if open source software (OSS) leads to businesses that by nature pursue a product-led growth (PLG) model. And we won’t take my word for it.
Instead, we’ll hear from Hashicorp CEO Dave McJannet, whom we spoke with after his company reported earnings last week. And we’ll leverage data regarding venture trends in the OSS market from the aptly named OSS Capital. Let’s have some Friday fun!
OSS as PLG
Hashicorp’s most recent earnings report beat market expectations, something that is a bit less common this year than the last. Regardless, what we care about is that Hashicorp is performing better than anticipated despite a somewhat slack macroeconomic picture. That’s a credit to the company, sure, but also its model, which is holding up despite changing economic winds.
Let’s talk about the model. We’ll start with McJannet’s perspective on how OSS leads to standardization thanks to network effects associated with open development (quotes edited for clarity and length):
I think open source is a very unique model. We don’t necessarily think about trying to do the [free to paid] conversion, necessarily. Our thesis is [that] by making our products open source, we can help drive standardization for everybody. [Our infrastructure product] Terraform [is a] really good example. You can use Terraform to provision your Amazon compute, your Azure compute, your Google compute, your Oracle compute. There are actually over 2,000 providers available, because the market is standardized. And that’s all available completely on open source.
Our goal with open source is actually to drive standardization for users — we think it’s good for everybody. Take the Linux example. There’s a reason Red Hat Linux became the de facto standard — because at some point all the hardware vendors built drivers for Red Hat. They didn’t do it for other distributions; they built for one. And that’s ultimately good for everybody. It means you can now safely use Red Hat Linux on any piece of hardware [just as] you can now use Terraform for anything for any type of [cloud] provisioning.
If you can build an open source tool that reaches something akin to ubiquity, you will have produced something that has its own pull; because it is the standard, it will attract usage simply by existing. Naturally, this is a slightly infrastructure-specific perspective on OSS and why it can generate its own in-market gravity.
But I would posit that, in effect, the same concept applies in micro for other open source products even if they lack an infrastructure focus. My argument is that open source code is simply easier to vet and adopt than closed source commercial code for many. In turn, that ease creates more usage, and, therefore, more of the same standardization that McJannet discusses above, even for other types of software.
Now, for the second part of the OSS equation, turning the potential in-market ubiquity of open source code into money:
So [the pursuit of standardization] is really step one of what we’ve done, [which] requires a very long-term view. And then, when people start adopting cloud, they start adopting open source products in their [version one] cloud journey. But then when an organization says, “Hold on a second, I need to apply some policy and governance to what’s being provisioned with Terraform,“ that’s what our commercial products do. So there’s a very clear delineation between what we do — one is for the use of practitioners and users, one is for the use of organizations orchestrating the use [of Hashicorp open source code].
It’s not a conversion process we’re trying to do, it’s actually a very different exercise. It’s, “Hey, you’re using Terraform. Trust me everywhere — let us come and talk to you about the value proposition of cost savings, [of] risk reduction associated with using Terraform but in a managed way.” So we don’t really think about that direct conversion as much as we do about engaging every enterprise on the globe, because everybody has this problem. So it’s much more of a direct conversation than a pure conversion.
In summary, OSS code tries to get everyone to use it and help guide its direction. Then the commercial conversation follows. The product itself leads to financial growth, if you will. That’s what most folks call product-led growth, or PLG.
PLG matters more than most other startup models today because venture capital demands are changing. As investors hunt for startups with more efficient growth, PLG can shine. Thanks to its ability to drive more sales from existing customers as their usage rises — the product leading the growth — it can help foment companies that spend less on sales and marketing. In turn, that limits cash burn, which investors today covet.
My perspective after yammering with McJannet and tracking the changing face of venture demands in terms of startup performance is that the inherently PLG-esque OSS model worked before the market changed but is even better set up to succeed today, at least from a fundraising perspective.
For startups, that’s something worth keeping in mind.
Is the money following?
If our little thesis is correct, we would expect to see more venture dollars flowing into open source startups over time. Per OSS Capital, such startups raised around $13.2 billion in 2021. Through August — the dataset in question was shared during that month, so we can presume that the final tabulated month is partial — the figure is $7.2 billion. That puts OSS startups on pace for around $10 billion in fundraising this year, or a bit more if we presume that the July and August data that OSS Capital counted will fill in a bit more over time, as venture capital data tends to.
If capital invested into OSS startups is set to drop modestly this year, is that indication that our thesis is incorrect? The opposite, actually. Recall that venture totals for nearly every sector and geography are declining this year; other sectors are seeing steeper declines, making projected OSS fundraising tallies for 2022 actually somewhat bullish in comparison.
The ever-present risk when connecting disparate trends is that we’re over-indexing on our own cleverness. But in this case, our argument is simple. OSS appears to lend itself to PLG. And PLG is increasingly popular, as OSS itself becomes ever-more accepted. The union of those two trends should be a few more IPOs when the market for public debuts opens up again. And more than a few venture rounds in the coming quarters. We’ll be watching.